AI Engineer - No-Code Platform

About the role

An AI-driven no-code platform is being built that turns ideas, prompts and designs into production-ready web and mobile applications. The AI Engineer designs and ships the intelligence behind it: systems that understand user needs and generate working UI, backend logic and databases. The role also covers researching and building alternatives to frontier model APIs, including self-hosted and on-premise LLMs. The position is onsite in DHA Phase 3, Lahore, with a 1 PM to 10 PM schedule and requires 2-4 years of experience.

Key responsibilities

• Design and build LLM-powered pipelines that turn natural-language requirements into application code, data models and APIs.
• Develop multi-step AI agents for planning, code generation, self-correction and automated bug fixing.
• Build design-to-code features (Figma, XD, wireframes) using vision and multimodal models.
• Engineer prompts, context management and retrieval (RAG) so outputs follow the component library and coding standards.
• Create evaluation frameworks for generated code quality, accuracy and reliability.
• Optimize model latency, cost and token usage across providers (OpenAI, Anthropic, open-source).
• Work with frontend, backend and product teams to integrate AI features into the builder.
• Evaluate open-source LLMs (Llama, Qwen, Mistral, DeepSeek) and benchmark self-hosted models against frontier ones.
• Design on-premise and private-cloud LLM deployments and explore optimization techniques such as quantization, distillation and fine-tuning (LoRA/QLoRA).
• Prototype hybrid routing between small in-house models and frontier models, and prototype new ideas.

Required profile

• Strong Python, Go and/or TypeScript skills.
• Hands-on experience building production applications with LLM APIs.
• Experience with agent frameworks (LangChain, LangGraph, LlamaIndex or similar) and vector databases.
• Experience with AST parsing, code analysis or compilers.
• Experience serving open-source LLMs with tools like vLLM or Ollama.
• Understanding of GPU infrastructure, model memory requirements and inference scaling.
• Experience with structured outputs, function/tool calling and code generation.
• Nice to have: fine-tuning or evaluating open-source models, prior work on no-code or developer tools, knowledge of MCP or tool-integration protocols.

Required skills

• Cloud deployment (AWS or Azure) and Docker.
• Structured outputs, function/tool calling and code generation.